Striping provides no redundancy at all: every file is spread across every member. If one disk fails, a slice of nearly everything will be lost. Recoverable; however odds are based upon condition of failed disk.
RAID 0 does not keep any redundant copy, therefore recovery means reading all members including failed, then assembling stripe from images. Amounts that can be recovered are dependent on ability to read the damaged disk.
The RAID 0 configuration writes multiple pieces of data across all drives in rotation, which is what provides speed and also is what causes fragility. With no parity and no mirror, there is nothing that can be computed back when data is lost. When one drive fails, several pieces of data are lost; nearly all files will have some portion missing or corrupted, which is why the volume is unable to mount instead of continuing to degrade. Two drives in a RAID 0 system fail at approximately double the rate of one drive, and the effect of losing either drive results in complete failure of the system.
All members (including failed ones) are imaged read-only after any physical repair they need (i.e., donor head replacement; PCB repair; firmware update/flash to obtain enumeration). Then, using the actual data within those images: block size, disk order & start offset are all derived based on trial-and-error until we can assemble a logical file structure out of what was previously recovered. From the images we create a virtual volume & extract the files from that.
When the casualty images completely, full recoveries generally occur. When portions of the casualty are unreadable, we find the missing blocks dispersed throughout the volume & the impact varies by file type: larger files are most frequently affected & container types that support gap-tolerance (some video/image files) may still function but w/ defects, while databases/archives will rarely ever work. For this reason, we always identify by name those files impacted by damage (as opposed to simply delivering them back silently mangled). This is the difference between “recovering” a file set vs. “counting” files.
Shut the computer down rather than trying again. Don’t let a controller or the operating system write new data on top of the old in an attempt to “recreate” or “initialise” the array, as the old data will be necessary to determine the correct configuration. Number each hard drive with its corresponding bay number when you remove them so that they can be put back into their proper positions during re-configuration. Do not turn on a failing disk until we have received it and are able to diagnose what needs to be done.
Yes, usually. RAID 0 doesn’t have redundancy to protect against failure. The outcome will follow the status of the failed disk. Each of the disks is imaged (read only), with the failed one physically repaired first if necessary, then we reconstruct the striped set from these images. If we recover the complete data on an individual disk, chances are you’ll get the complete recovery.
Because RAID 0 stripes your files across all of its disks in sequential fashion. When you lose a single disk you’re missing a piece of almost every file. Therefore, there’s no way to completely read the entire file system. This is what RAID 0 loses for fast write speeds and high capacity: zero fault-tolerant capabilities.
All of them. Every file is split into multiple pieces on each disk. Without each and every member disk, you can’t rebuild the stripe. Label them by bay order before removing them, as disk order matters for the reconstruction.
The price will depend on what has gone wrong. We charge from £500 +VAT for RAID, NAS and server work. This is fixed in a written quote after a free 48 hour no obligation diagnostic. When the failed member needs physical work to the drives, a 50% deposit is required before starting the repair, and the rest will only be charged when we have successfully recovered your data.
No. This will write new metadata to the disks, potentially destroying the very same structures needed to recover the original stripe geometry. So in short, leave the disks where they are once you discover that your RAID has failed. Send them in as they were when you discovered the problem.